GWTF

Industrial engineer turned AI tech lead · problem-solver · founder of GWTF (go with the flow) · Lower Bavaria, Germany (greater Munich area)

Florian Wasmeier

Optimistic realist and curious problem-solver. I don't really build software — I solve problems, and AI is the sharpest tool I've ever been handed. Go with the flow.

I grew up in the countryside of Lower Bavaria — a peaceful place full of trees, gardens, birds and family. As kids we practically lived in the forest: we built our own little village out there, mined stones, invented our own currency and a whole trading system. At the same time we were playing The Settlers on the PC, and the two blurred together — building and trading economies in a game, building and trading them for real in the woods. That's the earliest version of me there is: someone who wants to build a system, understand how it works underneath, and then optimise it to the last percent.

That thread never left. Strategy and builder games — Age of Empires, Civilization, Anno — where I obsess over finding the perfect setup. Minecraft, but the heavy tech modpacks with engines, refineries and reactors, where you actually have to read documentation to build a machine. Taking apart old PCs in the basement with a neighbour just to see if we could get them to boot again. I was a talented but lazy student — logical thinking, physics and maths came easily, and I never learned by sitting in lectures. I learn from an inner drive, and that drive is curiosity: I need to understand how things work, what's really going on beneath a problem, a process, or a person.

I studied industrial engineering (Deggendorf, then a master's at FAU) — honestly, partly because I couldn't decide what to be and it kept every door open. It turned out to be exactly right, though I only understand that now. My first real programming was a C++ internship at the Fraunhofer Institute, controlling machines and APIs from code — and it clicked instantly. Programming felt like Minecraft: solving fascinating problems, building things. A semester abroad in South Korea taught me what real determination looks like — people driven by decision, not by mood. My master's thesis put me into machine learning and neural networks, predicting geometry from point-cloud vehicle data, and I knew: AI is a once-in-a-generation opportunity for humanity, and I want to be part of it.

Somewhere along the way I realised I'm not actually an "IT person." There isn't one single thing I want to do. What I love is solving hard, complex, systemic problems — and IT, and now AI, are simply the most powerful tools for that. The breadth of industrial engineering — thermodynamics, business, lean, HR, org development — used to feel like being a master of nothing. With AI in hand it became my superpower: I can talk to anyone in any department, understand what they actually do, and aim AI at the real problem.

GWTF — go with the flow — is my studio and my passion. "Flow" is literally in Florian; "go with the flow" is how I solve problems and how I want to live — the flow state from Ikigai, creative problem-solving, having fun, sometimes joking about it. I started it last Christmas and pour 20–30 hours a week into it on top of a full-time job, because I want to show people what's possible: that with AI you get superpowers, that you can take your own ideas and actually build them. Everybody can follow their dreams. That's the whole point.

Highlights

  • Founded GWTF (go with the flow) — an independent studio built on three pillars: Studio (consulting & product work), Products (things I build), and Culture (growth, curiosity, flow).
  • Building Grove — a calm, cozy app for reading books, reflecting on their ideas, and understanding yourself in a fast, messy world.
  • Building SimplyStocks — a beginner-friendly, radically simple overview of what a company actually does and how it makes money, so people can invest in things they understand.
  • Building Atelier — an AI image-generation studio that lets fashion/apparel brands visualise how their clothing looks on models without a real shoot.
  • Built the GWTF site itself, including an MCP 'interview me' tool and a personal access-key system for a deeper, key-holder-only profile.
  • Master's thesis: machine-learning models predicting point-cloud geometry for automotive inspection — reframed a 15M-point-per-vehicle problem into small ~10k-point 'islands' around each feature, and improvised a 5-PC compute cluster from scavenged machines to cut dataset extraction from ~3 months to ~3 weeks.
  • Leads the AI function at an industrial company: enablement and training across technical and non-technical staff, product-portfolio and security decisions, prototyping agentic/MCP systems, evaluation & tracing, RAG, and the company's AI vision — managing a small technical team.

What I work in

Product sense & taste — knowing what to build and where the real problem is · AI system design — agent harnesses, RAG, evaluation & tracing, context engineering · Context engineering — the right context, tools and skills present at the right step (and discarding the rest) · Problem & product discovery — hunting the root problem beneath the symptom · Systems & software architecture — the big-picture structure, not the nitty-gritty · Full-stack development — frontend, backend, databases, infrastructure · Security-first engineering — RBAC, row-level security, authentication, end-to-end in production · Machine learning — neural networks, point-cloud / spatial data · Cloud & infra — Kubernetes, deploying and testing applications · Cross-functional communication — from shop floor to C-level, plus universities and external partners · Workshop facilitation — discovery, prototyping and vision-building sessions with departments · Systems thinking applied to himself — habits, focus, and getting into flow

How I think

I build Autobahnen, not country roads. When someone brings me a problem, I don't just want to solve it — I want to know what's lying beneath it. Maybe we're only fixing a symptom; maybe there's a root problem, and fixing that kills other problems we didn't even know existed yet. So wherever it fits, I want discovery first: how the processes really work, how the problems actually occur, what each person expects — and then we design the system underneath.

AI is just another building block. Code is a deterministic outcome; a large language model is a statistical one — and we want to reduce that statistical surface wherever we can. That's what agent harnesses are for: a controlled environment around the AI, where you decide exactly what goes in and what comes out. Context engineering matters just as much — which context is present at which step, because you're optimising the likelihood of the best outcome at that moment.

Security is where I overspend on purpose; MVPs are where I cut corners. RBAC, row-level security, real authentication, secure end-to-end in production — non-negotiable. Everything else, I iterate fast, especially UX, because UX is where taste now lives. Front-end is becoming commoditised by AI; the interfaces that matter are increasingly chat, then voice, with visualisation where it earns its place.

Architecture is the human's job. The nitty-gritty of writing code is exactly what AI is good at. The big picture — do we need a cache, which database, does this even need to be AI at all — that's where human intelligence and taste come in.

Action comes before motivation. Motivation doesn't fall out of the sky; it comes from action, and action feeds more action. You decide for it every morning — not yesterday, not tomorrow. That's what I mean by driven. Mind over thoughts: thoughts come and go, but I get to reason about them and choose what to do with them. And ego is the death of creativity — take the work seriously, never yourself.

I'm an optimistic realist: optimist because I believe we can solve every problem, realist because I know the actual constraints and what the technology can and can't do.

Outside the work

Sports of every kind — running, swimming, lifting, yoga, and a background in soccer, tennis and table tennis; yoga doubles as a way to reason about my own thoughts · Surfing and surf culture — the vibe is a big part of what 'flow' means to me · Travel and cultures — food, music, dance, history, the different energy of a place; big on Okinawa and the blue zones · Strategy & builder video games — Civilization, Anno, Age of Empires; Minecraft tech modpacks; and competitive League of Legends (peaked Master) · Music — genre-hopping constantly; music as a flow state you don't have to be productive in · Investing & financial freedom — and making it understandable for people who find it scary · Reading — self-improvement, psychology, economics, philosophy; I read to build myself · Food & nutrition — vegetarian, with a real interest in health and longevity · Animals and nature — grew up surrounded by both and never lost it · Nerding out about sleep — REM, circadian rhythm, why dolphins sleep with half a brain

A few questions

What do you actually do?

I solve problems. For a long time I'd have said 'I build software,' but that's not really it — software and now AI are just the most powerful tools I've found for the thing I actually love, which is untangling hard, complex, systemic problems. These days my value is much more in product sense, taste and architecture than in typing out code, because AI does most of the typing.

What is GWTF / 'go with the flow'?

It's my studio and my passion project. 'Flow' is in my name — Florian — and 'go with the flow' is how I work and how I want to live: getting into the flow state, creative problem-solving, having fun with it. It has three pillars: Studio (consulting and product work for others), Products (things I build, like Grove), and Culture (growth, curiosity, travel, music, self-improvement). Underneath it all I just want to show people that with AI you can actually build your ideas.

What are you building right now?

Three things. Grove — a calm place to read books, reflect on them, and understand yourself in a world that never slows down. SimplyStocks — a dead-simple overview of what a company really does and how it makes money, because you shouldn't invest in things you don't understand. And Atelier — an AI image-generation studio so fashion brands can see their clothing on models without a full shoot. Every one of them started as a problem I had myself.

How do you use AI in your own work?

As a building block, not a magic box. Code is deterministic, an LLM is statistical, and I try to shrink that statistical surface wherever I can — controlled agent harnesses, careful context engineering, evaluation and tracing. I let AI do the nitty-gritty coding and spend my own attention on the architecture and the taste, which is where humans still win. And I'm relentlessly efficient about it — I'll have parallel coding sessions and a research agent running while I'm doing something else entirely.

Where does the name 'Grove' come from?

It's a bit of a wordplay. I'm German, the app is all about growth — and if you're not great with the 'th', 'growth' comes out as 'grove.' It also just fits: Grove is meant to be a cozy, calm place that holds your most inner reflections. So it's fitting and a little funny at the same time, which I liked.

Do you take on client work?

Some, through the Studio — but weighted toward product sense, discovery and taste rather than me being a pair of hands, because AI already handles most of the actual building. What I'm most drawn to is advisory and inspiration-giver / challenging-coach roles, co-founding or collaborating on products, and honestly just genuinely curious conversations. Make me curious and there's a very good chance I'll want to talk.

Go deeper

Want the fuller picture — my full CV, candid answers, and how to reach me? That sits behind a personal key. If I've given you one, open the dossier.

Prefer to let your own AI do the asking? Point it at my MCP server mcp.gowiththeflow.io and call the interview tool — it'll answer as me from this same background. With a key, it unlocks the deeper material too.